Home / Strategies
The wall of verdicts

Honest strategy verdicts for every asset

We tested ~382 indicators per asset across 903 assets — 660,005 backtests — then ranked every setup out-of-sample and applied a multiple-testing correction, because picking the best of hundreds of tries is data-mining by construction. Most assets fail. We publish that loudly: it's the whole point.

10
validated after correction · 1% of assets
190
mixed — can't rule out luck · 21% of assets
703
nothing beat buy & hold · 78% of assets
Why so few pass

VALIDATED requires beating buy-and-hold in the full window AND out-of-sample AND an out-of-sample Sharpe above the selection hurdle √(2 ln N)/√T. MIXED beat buy-and-hold consistently but not by enough to rule out being the luckiest of ~760 tries. Everything else is NOTHING BEAT BUY-AND-HOLD — the most common and most useful verdict on this site. Every page is forward-tracked in public from publication. Not investment advice.

AssetVerdictTop setup (OOS-ranked)TF OOS Sharpe ↕ Hurdle OOS α ↕

Hypothetical backtests, costs included. Verdicts are statistical judgments about the past, not predictions. See each page's methodology box and the disclaimer.